Ann. Occup. Hyg., Vol. 54, No. 8, pp. 923–933, 2010
Ó The Author 2010. Published by Oxford University Press
on behalf of the British Occupational Hygiene Society
doi:10.1093/annhyg/meq063
Measurement of Action Forces and Posture to
Determine the Lumbar Load of Healthcare Workers
During Care Activities with Patient Transfers
ANDREAS THEILMEIER*, CLAUS JORDAN, ALWIN LUTTMANN and
MATTHIAS JÄGER
Leibniz Research Centre for Working Environment and Human Factors (IfADo), Ardeystraße 67,
44139 Dortmund, Germany
Received 15 October 2009; in final form 11 July 2010; published online 17 September 2010
Moving patients or other care activities with manual patient handling is characterized by high
mechanical load on the lumbar spine of healthcare workers (HCWs). During the patient transfer activity, the caregivers exert lifting, pulling, and pushing forces varying over time with respect to amplitude and direction. Furthermore, the caregivers distinctly change their posture
and frequently obtain postures asymmetrical to the median sagittal plane, including lateral
bending and turning the trunk. This paper describes a procedure to determine lumbar load
during patient transfer supported by measurement techniques and an exemplary application;
this methodology represents the basis of a complex research project, the third ‘Dortmund
Lumbar Load Study (DOLLY 3)’. Lumbar load was determined by simulation calculations using a comprehensive biomechanical model (‘The Dortmunder’). As the main influencing factors, the hand forces of the caregiver exerted during typical patient transfers and the posture
and movements of the HCW were recorded in laboratory studies. The action forces were determined three-dimensionally with the help of a newly developed ‘measuring bed’, two different ‘measuring chairs’, a ‘measuring bathtub’, and a ‘measuring floor’. To capture the forces
during transfers in or at the bed, a common hospital bed was equipped with an additional
framework, which is attached to the bedstead and connected to the bedspring frame via
three-axial force sensors at the four corners. The other measuring systems were constructed
similarly. Body movements were recorded using three-dimensional optoelectronic recording
tools and video recordings. The posture and force data served as input data for the quantification of various lumbar-load indicators.
Keywords: action forces; back pain; biomechanical model; health-care workers; lumbar load; manual handling;
nurses; posture
are one of the most frequent causes for healthrelated absenteeism in the workplace (European
Communities, 2002; BKK-Bundesverband der
Betriebskrankenkassen, 2008). Similarly, occupational low back pain is a significant problem among
nurses (Mitchell et al., 2009) and care activities with
patient transfer may lead to high load on the spine
and may accelerate the development of degenerative
disc-related diseases in the long run of occupational
life (Videman et al., 2005).
At the Leibniz Research Centre for Working Environment and Human Factors at Dortmund University
INTRODUCTION
Manual materials handling activities are connected
with a high risk for the development of diseases
related to the intervertebral discs (Videman et al.,
1984; Luttmann et al., 1988; Riihimäki et al.,
1989; Hofmann et al., 1995; Hofmann and Korn,
2001; Seidler et al., 2009, 2003). Furthermore,
diseases of the muscle and skeleton systems
*Author to whom correspondence should be addressed.
e-mail: [email protected]
923
924
A. Theilmeier et al.
of Technology (IfADo), in cooperation with the Statutory Accident and Health Insurance Institution for
Health Services and Welfare Care (BGW), the
‘Dortmund Lumbar Load Study 3 (DOLLY 3)’ was
conducted (Jäger et al., 2007; Theilmeier et al.,
2010). The study represents a long-time research
project on the determination of lumbar load in selected care activities with patient transfers. It was accomplished to assess lumbar diseases with respect to
mechanical loads in occupational fields and for the
development of preventive procedures for the avoidance of lumbar spine’s disease. This paper describes
the first part of the study and is primarily concerned
with the methodology to determine the lumbar load
of healthcare workers (HCWs) performing patient
transfers supported by measurement techniques and
an exemplary application.
When moving patients during care activities,
HCWs frequently adopt postures asymmetrical to
the median sagittal plane, including lateral bending
and turning the trunk, and laterally positioned arms
including sideward force exertion. Furthermore, they
exert lifting, pulling, and pushing forces varying
over time with respect to amplitude and direction.
Therefore, a three-dimensional (3D) determination
and replication of both, the posture and the action
forces in high temporal resolution, are needed for
an adequate approach for quantifying biomechanical
indicators of load on the lumbar spine (Jordan et al.,
2006; Theilmeier, 2006).
In the research project, the quantification of the
lumbar load was performed by model calculations
using a previously developed simulation tool (‘The
Dortmunder’; Jäger et al., 2001a). The calculation
of mechanical lumbar-load indicators with this 3D
multisegmental dynamic biomechanical model presupposes knowledge; firstly, on forces acting on
the nursing person’s body, commonly applied at
the hands (so-called action forces) and, secondly,
of caregiver’s posture.
With a similar approach in former studies, Garg
et al. (1991) as well as Owen et al. (1992) examined
the biomechanical load on the lumbar spine with
a static biomechanical model, whereby they assumed that the forces at the hands of the HCW correspond to the half patient weight. Furthermore, in
their model, lumbar-load computations were only
possible for vertical action forces, i.e. forces acting
spatially could not be considered. In a similar way,
Zweiling (1996) accomplished the biomechanical
computations with a comparatively simply structured ‘three-angle model’. This permits analyses of
symmetrical slow lifting operations in the sagittal
plane only. By recording vertical forces with mea-
suring soles under the feet of the HCW, Morlock
et al. examined care activities; in consequence of
this procedure, horizontal forces on the worker
could not be considered (Morlock et al., 1997).
Deuretzbacher and Rehder (1997) used a biomechanical model, which represents the musculature in the
lower back by two force resultants acting in parallel
to the spine. This very limited muscular structure
modelling permits adequate computations for activities without lateral force components only. Because
of the limitations of the described methods, the new
technology—described in this paper—was developed aiming for a more detailed and close-to-reality
description of lumbar load during patient transfer
activities.
METHODS
The posture and the forces exerted by the HCW
are the most relevant factors concerning the lumbar
load of a person. The data are needed as input to calculate the lumbar load with the biomechanical
model. Following the aim of a detailed determination of the lumbar load presupposes knowledge of
the time courses of the action forces and the posture
3D and in high time resolution. To this end, posture
and force data were recorded with a sample rate of
100 Hz. The methodology used for the studies requires high personnel effort for gathering and evaluation of the data, and therefore, it is less suited for
larger groups of subjects. The disadvantages of evaluating a relatively low number of task executions
could be reduced by a careful selection of typical,
i.e. ‘average’ tasks from the pool of the totally collected data and by utilizing the help of two highly experienced caregivers acting alternately as a ‘typical’
HCW or a typical patient.
Posture recording
The posture and the movements of the nursing personnel were gathered with the help of a combination
of two measuring systems: video analysis and optoelectronic measurements. Figure 1 gives an overview
of the laboratory and the equipment.
The posture was captured via video documentation from different lines of vision with four video
cameras (Video 1 up to Video 4). Additionally, a
3D optoelectronic motion capturing system was used
to track continuously the coordinates of small infrared markers (diameter 10 mm, active area ,1 mm)
attached to relevant body parts of the subject (shoulder, hand, hip, and heel, each at the right- and lefthand side) and, for reference, at the bed frame.
The markers’ spatial localizations were recorded
Measurement of action forces and posture to determine the lumbar load of HCWs
925
Fig. 1. Drawing of the laboratory with top view of the measuring area (grey) and details regarding the two combined OPTOTRAK
position sensors and the four video cameras.
via two ‘position sensors’ consisting of three infrared cameras each. The position sensors were
mounted at opposite walls of the laboratory. The
measuring space of the position sensors overlaps
and so a common measuring area is formed enabling
the tracking of the markers at both sides of the body
of the subject. The aperture angle of the optical systems leads to a hexagonal measuring area in the top
view (coloured grey in Fig. 1; length 5.0 m, width
between 1.2 and 1.6 m, and height 2.0 m). The
position sensors act cooperatively and, therefore,
a changeover of a marker from the measuring area
of the first sensor to the measuring area of the second
sensor was possible. So failures in the detection of
the markers (e.g. due to covering of the markers by
the patient) could be significantly reduced.
Applying a graphical animation system, the real
posture of the nurse was reproduced digitally in an
iterative procedure by combining the video and the
optoelectronic data. Posture data were transformed
into a digital format suitable for the subsequent
lumbar-load quantification via applying the computerized simulation tool The Dortmunder. For the complete posture replication of the observed person, the
data of the angular position of each body segment
were reproduced on the basis of the video recordings
and a manikin was presented using a stick figure animation. With specific software, the 3D coordinates
of each joint and body segment were determined in
a body-related coordinate system. In a recursive procedure, the coordinates were changed and the position and movement of the stick figure were adapted
to the position and movement of the infrared markers
(Jordan et al., 2003).
Determination of action forces
With the help of newly developed devices—a
measuring bed, different chairs, bathtub, and
floor—the caregiver’s action forces applied to the
patient during a transfer were captured continuously
during the transfer. The forces were not measured
‘directly’ at the hands of the caregiver, but ‘indirectly’ by the change of the forces transferred to
the measuring device, i.e. the change in the reaction
forces of the measuring bed etc., is considered to figure the action forces applied by the nurse. The forces
had to be captured regarding amplitude, direction,
and the point of application. When configuring the
measuring systems, special emphasis was put on
copying the structure and the functionality of the underlying devices as exactly as possible (Theilmeier
et al., 2003, 2006).
For the construction of the measuring bed (see
Fig. 2, left-hand part), a common hospital bed was
equipped with an additional framework attached to
the bedstead and connected to the bedspring frame
via triaxial piezo-ceramic force sensors at the four
bed corners. To prevent the bending of the bedspring
frame, which results in malfunction of the force
sensors, the bed was stabilized by inserting an
additional frame structures and several supporting
stands. Additionally, two ‘measuring bars’, each
equipped with two triaxial force sensors, were
mounted on the bed—one at the bed’s long side,
the other at the bed’s head—allowing the measurement of forces induced by a leaning of the HCW
against the bed. Two force platforms at the floor
were used to capture the ground reaction forces of
the HCW when the patient leaves the bed and so
926
A. Theilmeier et al.
Fig. 2. Left: measuring system ‘bed’ with original (upper) and additional (lower) bed framework and two force platforms as well
as an enlarged cut-out showing one of the four force sensors mounted at the bed corners; top right: Measuring system ‘chair’ with
three force platforms; and below right: measuring system ‘bathtub’ consisting of two force platforms and a steel framework.
the measuring function of the bed fails. To enhance
the contact of the force platforms to the laboratory
floor and to level surface unevenness, the platforms
were placed on concrete plates.
The basis for the determination of the action
forces is described representatively for all used force
measuring devices in the following for the measuring bed: The action–force amplitude and its direction, i.e. the ‘overall resultant’ action force result
from the vectorial sum of all force components measured with the triaxial force sensors at the four bed
corners—diminished by the force due to the patient’s
weight. For the calculation of the point of force application (distance between action force vector and
the reference point at the middle of the bed’s surface), the vectorial sum of moments could be drawn
with the applied moment, the measured moment, and
the moment caused by the patient’s weight (‘patient’s moment’). The applied moment contains the
overall resultant measured action force and the associated point of force application that is searched. To
determine the patient’s moment at each point in time
during the transfer, the patient’s centre of gravity
shift was gathered with the help of accompanying biomechanical model calculations considering patient’s
movement during the transfer.
The combination of the measuring bed, the measuring bar at the bed’s long side, and the force platforms, furthermore allow the calculation of the force
distribution when two HCWs handle a patient at the
same time. The action force of the first caregiver can
be calculated from the ground reaction forces diminished by the forces measured with the bar in due
consideration of the respective point of force appli-
cation. The action force of the second HCW is determined from the force measured with the bed
diminished by the action force of the first HCW.
To examine transfer activities to a chair or from
a chair, the system ‘measuring chair’ was developed
on the basis of a commonly used patient chair (see
Fig. 2, right-hand part, above). In a first version of
the measuring chair, the patient chair was mounted
on a single force platform with four 3D force sensors; it was used to gather the action forces of the
HCW during simple transfer activities like ‘raising
a patient from sitting to upright standing position’.
For more complex patient movements, the system
was modified applying two additional force platforms (see Fig. 2, right-hand part, above). Determining action forces with the system, measuring chair
works according to the same principle as described
above for the measuring bed.
Due to the dimensions and characteristics of the
patient chair (e.g. existence of lowerable arm rests
or material of the upholstered seat), it was possible
to reproduce many actions commonly performed in
the care sector like ‘placing a patient from sitting
at bed’s edge in a chair and vice versa’ or ‘raising
a patient from sitting to upright standing position
and vice versa’ with high authenticity. The height
of the measuring chair can be adapted according to
the respective requirements. Since the forces transferred by the HCW to the patient should be measured, a contact of the caregiver to the force
platform was avoided by using two footstep bridges
positioned above the platform.
For the determination of the nurse’s action forces
during the activity ‘moving the patient into the
Measurement of action forces and posture to determine the lumbar load of HCWs
bathtub’, the system ‘measuring bathtub’ (see Fig. 2,
right-hand part, below) was developed. In order to
simulate the care situation close to reality, the dimensions and geometry of the measuring bathtub
were arranged according to a commonly used bathtub. For this purpose, a usual bathtub was installed
onto a steel framework. Forces, applied on this construction, were measured with the help of two force
platforms between the framework and the laboratory
floor. The contact of the framework to the force measuring platforms was established using four contact
elements adjustable in height in order to guarantee
a uniform force transmission. The height of the
standing area of the caregiver could be adjusted using pedestals of various heights.
For the analysis of patient transfers on the floor,
like ‘raising a patient from sitting on the floor to sitting or to upright standing position’, the nurse’s action forces were measured using the measuring
system ‘floor’. With the help of two force platforms,
arranged side by side, the HCWs’ action forces were
captured indirectly by the measurement of the
changes of the ground reaction forces. The adaptation of the standing area of the HCW to the overall
height of the force platforms and concrete plates
was obtained by using several pedestals. Since the
friction between the patient and the metallic surface
of the force platform was too small in comparison to
real conditions, so-called anti-slide mats were positioned on the force platforms. By this measure, an effective support of the HCW was achieved and close
to reality conditions were established.
The lumbar load of the HCW was determined by
model calculations applying the 3D dynamic simulation software tool The Dortmunder. This tool permits
the quantification of several low back load indicators
considering the gravitational and inertial effects of
the body and a handled object or—here—a subject,
the effects of asymmetry of posture, movement,
and force exertion, as well as the effects of intraabdominal pressure in supporting the trunk during
forward-inclined positions. The lumbar load is described by The Dortmunder on the basis of mechanical characteristics, like compressive or shear forces
and bending or torsional moments, with respect to
the intervertebral discs in the lumbar section between the first lumbar vertebra (L1) and the upper
part of the sacrum (S1). The calculations in this
paper concern the lowest intervertebral disc in the
spine called ‘L5–S1’ as reference point. The lower
lumbar spine is generally accepted as a ‘bottleneck’
since it represents the structure with very high
risk for the development of lumbar degenerative
diseases.
927
TYPICAL APPLICATION
The measuring systems were developed in order to
study the lumbar load of HCWs during manual patient handling actions, which are presumably accompanied with high biomechanical load on the lumbar
spine. In the following, exemplary measurements
and calculations regarding the activity ‘lifting a leg
of a lying patient and vice versa’ will be demonstrated. In Fig. 3, the posture of the nursing person
is represented at three basic points in time (postures
B, C, and E as described in the following) on the basis of three photos.
To obtain input data for the biomechanical model
calculations, the posture during the transfer activity
was reproduced with the help of the optoelectronic
data and by video recordings as described in Posture
recording.
The activity was divided into seven segments
(marked with the characters A–F) because with the
biomechanical model only unidirectional segment
movements can be analysed within one trial. The
segments are represented in Fig. 4 as stick figures
in side view: nurse standing in an upright position
(A), bending the trunk (A–B), lifting the patient’s
leg (B–C), holding the leg (C–D), lowering the leg
(D–E), nurse straightening up (E–F), and standing
in upright position again (F). The activity was partly
performed by the HCW with clearly inclined trunk
(positions B and E in Fig. 3). In addition, for the execution of the activity, the nurse adopted a slightly
side-bent posture (positions B and E in Fig. 3).
Represented in Fig. 5, time courses of the horizontal (forward/backward and leftward/rightward) and
vertical (upward/downward) action force components were recorded for the activity lifting a leg of
a lying patient or vice versa. The segments are
marked with the same characters as the stick figures.
In the sections ‘HCW upright’, the HCW was in
the so-called basic position without applying any action forces. In the second section, the caregiver bent
down to the patient and—in order to grasp her leg to
raise—supports herself on the bed. Resulting from
that at the end of the section, a force of 70 N rightward and up to 50 N in vertical direction (downward)
was determined. The third section ‘lifting patient’s
leg’ is characterized by a steep raising of the upward
vertical force component up to 90 N and a horizontal
lateral force component changing from 70 N rightward to 30 N leftward. During the section ‘holding
patient’s leg’, the caregiver held her leg using an upward force of 75 N and exercised an easy force backward to the patient. In the next section ‘lowering
patient’s leg’, the vertical force component declines
928
A. Theilmeier et al.
Fig. 3. Video prints of the care activity ‘lifting a leg of a lying patient and vice versa’ in conventional execution with a passive
patient.
Fig. 4. Stick-figures of the care activity ‘lifting a leg of a lying patient and vice versa’ in conventional execution with a passive
patient.
Fig. 5. Time courses of the components of the action forces of the HCW during the care activity ‘lifting a leg of a lying patient and
vice versa’ in conventional execution with a passive patient.
to zero when the leg lays on the bed surface and becomes negative, as soon as the caregiver supports
herself on the bed. The activity is finalized by the
sections ‘straightening up’ and ‘HCW upright’ with
action forces similar to the corresponding sections at
the beginning of the activity.
Measurement of action forces and posture to determine the lumbar load of HCWs
With the help of posture and force data, as indicators for the lumbar load, time courses of forces and
moments at the lumbosacral disc (L5–S1) were calculated by applying the biomechanical model. The
time courses of the force components at the lumbar
disc (shear force forward/backward, shear force
leftward/right, and the axial compressive force) are
presented in Fig. 6.
In the sections called HCW upright (posture A),
disc forces were determined, which are typical for
an upright standing person. The compressive force
amounted to 0.7 kN due to the weight of the caregiver’s body segments superior to L5–S1, the sagittal
‘forward’ shear force adds up to 0.2 kN due to anatomically caused tilt of the disc, a lateral ‘sideward’
shear force is missing because of the symmetrical
posture of the caregiver in this phase. In the second
section (the HCW bends forward to the patient, motion from posture A to posture B) due to only small
action forces applied by the caregiver, the typical influence of a motion with acceleration and deceleration shows up: the compressive force (0.7 kN) is
approximately as high as the value for the basic posture at the beginning of the section, is slightly increased to 1 kN to overcome the resting status,
rises then up to a maximum of 2.5 kN (maximum
deceleration when retarding the trunk in motion),
and drops finally to 2 kN (decreased deceleration
of the trunk to receive the intended inclined posture).
The compressive force achieved a maximum of
4 kN—the highest value for this activity—at the beginning of the third section (postures B–C) because
of a clearly forward bent posture of the caregiver
on the one hand and because of the acceleration of
both, the own moved body segments and the mass
of the patient’s leg on the other hand. During holding
929
patient’s leg (postures C–D), the disc compression
remains in a range of 3 kN. The two sections lowering patient’s leg and straighten up essentially correspond reversely to the process in the sections ‘lifting’
and ‘bending’. The time courses of the shear forces
follow the time course of the compressive force
approximately—however, with clearly smaller values
(highest values: 0.6 kN sagittal and 0.5 kN lateral).
The time courses of the shear forces follow thetime
course of the compressive force approximately—
however, with clearly smaller values (highest values:
0.6 kN sagittal and 0.5 kN lateral). The activity
described exemplarily above lifting or lowering
a patient’s leg was performed by two persons acting
alternately as a HCW or as patient (P1 and P2). The
time courses of lumbar-load indicators were analysed for totally nine executions were analysed (five
and four times for Patient 1 or 2, respectively).
Regarding the peak value in a time course of the
indicator ‘compressive force at L5–S1’, the average
amounts to 3.3 kN for P1 (range: 2.9–4.0 kN, n 5 5)
and 2.9 kN for P2 (2.6–3.5 kN, n 5 4).
DISCUSSION
Criticism of the methods
The methodology described in this paper is applied in a research project concerning the determination of indicators for the lumbar load of HCWs
during patient transfer activities. For this purpose,
measuring systems were implemented to determine
the main influencing factors on lumbar load, such
as the posture of the HCW and the forces transferred
to the patient. Prior to the laboratory measurements
simulating typical patient transfer activities in
Fig. 6. Time courses of the resulting lumbar load, i.e. of the forces transferred via the lowest intervertebral disc of the spine
(‘L5–S1’), for the care activity ‘lifting a leg of a lying patient and vice versa’ in conventional execution with a passive patient.
930
A. Theilmeier et al.
Table 1. Results of the inspection of measuring
accuracy—values in millimeter; set point 5 202 mm
n 5 100
Measuring condition
Parallel
45°
Mean value (mm)
202
202
Standard deviation (mm)
,1
,1
a hospital, quality characteristics of the measuring
systems were checked.
Posture recording. The optoelectronic posture
measuring system is equipped with two position sensors (see Fig. 1); each of them is assembled with
three infrared cameras. The accuracy of the position
detection of the infrared markers used to track the location and the movement of the nurse’s body segments depends on the adjustment of the cameras
within the position sensors. This adjustment was already performed by the manufacturer. Nevertheless,
after fixing the position sensors at the walls of the
laboratory, the overall accuracy of the total device
was checked.
For this purpose, two infrared markers were fixed
on a solid bar in a definite distance of 20 cm. The
bar was positioned at different places within the
measuring space of the laboratory (see Fig. 1) and
the distance between the markers was determined
using the total measuring device. The measurement
was performed using two different orientations of
the bar, one with the bar in parallel to the position
sensors and another with an angle of 45°. The results
of 100 measurements are shown in Table 1: the optoelectronically measured intermarker-distance corresponds to the ‘real’ distance measured with a
sliding rule and the standard deviation is , 1 mm,
i.e. ,1%, for both orientations of the bar.
The examination of the measuring accuracy substantiated that the used measuring system itself operates with a very high accuracy. For the measurement
of the position of HCW’s body segments, the overall
accuracy essentially results from additional factors,
such as the positioning of the markers distantly from
the real body joints on the subjects skin or clothing.
Some disadvantages of an optoelectronic system
usage—in particular, the occasional covering of the
markers by body segments of the HCW or the
patient—was compensated for by using the video
systems’ data. In such cases, the combination of
the optoelectronically measured data with the video
images allows to complete the position data for
relevant periods of time.
Determination of action forces. For the newly developed force measuring systems, comprehensive
testing and calibration procedures were performed.
In particular, the linearity, repetition accuracy, crosstalk, and the temporal drift for the entire force measurement systems consisting of the bed, chair etc.
with the integrated force sensors or force platforms
were determined. In the following, some characteristic findings are shown exemplarily, in particular for
the measuring bed. More detailed information and
results of the other measuring systems are provided
by Theilmeier (2006).
In several series, the force measurement systems
were charged with forces of different amplitude
and different directions. For the coefficient of determination (R2), a value of .0.99 was found in each
case indicating the high linearity of the force sensors, even if they are inserted in the measuring systems. By the use of a calibration in the relevant
measuring range, the absolute values for measurement errors were ,25 N in vertical direction and
,20 N in horizontal direction.
The reliability of the systems was examined on the
basis of in total 10 measurements for each direction.
A maximum deviation between the readouts of ,5%
was always observed. As to the crosstalk between
the signals representing the three spatial force components, values of ,3% were measured; for example, loading the measuring bed with a vertical
force of 850 N results in crosstalk signals in the horizontal directions of 25 N (3%, transverse bed axis)
and of 10 N (1%, longitudinal axis). Because the
used piezo-ceramic sensors exhibit an unavoidable
electrical drift, the temporal change in the readouts
was examined. The measured drift corresponds to
force changes of 0.3 N min 1 (if loaded with 270 N
vertically) and 0.5 N min 1 (unloaded). Since the
duration of the examined patient transfers is clearly
,1 min, the ‘drift influence’ can be regarded as
negligible.
A force applied to the measuring bed results in reaction forces at the four edges of the bed where the
force sensors are located; the readouts of the four
sensors depends directly on the localization of the
point of force application (PoF). Therefore, inaccuracies with the measurement of the forces, described
above, have a direct influence on the accuracy of the
localization of the point of force application whose
determination results from fixed geometrical dimensions of the measuring bed. Therefore, the verification of the point of force application accuracy is
studied exemplarily in detail. The bed’s surface is
shown schematically in Fig. 7 together with the insertion of 11 measuring grid points (closed circles)
where a vertical force of 150 N was applied successively in five series of measurement. An additional
weight of 68 kg, continuously placed close to the
Measurement of action forces and posture to determine the lumbar load of HCWs
931
Fig. 7. Verification of the point of force application (PoF) accuracy—measuring grid on the bed surface from 11 measuring grid
points and position of the computed points of force application (five series of measurements).
bed’s centre, simulated the patient (rhombus in Fig. 7).
As results of these measurements and the respective
computations, the calculated positions of the points
of force application are shown as open circles. The
deviation from target value (closed circle) to actual
value (open circle) amounted to 40 mm (mean
value 37 mm, minimum 15 mm, maximum 91 mm,
standard deviation 16 mm). Related to the measuring
range (2000 mm), the average deviation is ,2%, the
maximum deviation was ,5%.
The dynamic behaviour of the measuring device is
only of subordinated importance: For example, the
natural frequency of the measuring bed clearly exceeds 20 Hz, whereas a total patient handling lasts
few seconds with activity sections of about half a second. Nevertheless, the performed activities are to be
considered highly dynamic from the ergonomic
point of view; the influence of accelerations of patient’s and caregiver’s body masses including inertial
effects has therefore to be considered in lumbar-load
quantification, which is guaranteed, here, by applying the multi-linked 3-D dynamic validated simulation tool The Dortmunder.
Comparison with recent studies
The aim of the research project described here was
the estimation of lumbar load during care activities
on basis of measurement-supported posture and action force determination considering spatial and inertial influences. In spite of their doubtless merits
with other aspects in scientific analysis and prevention, previous biomechanically oriented studies into
nurse’s low back load can serve as a reference under
restriction only (Waters et al., 1993; Marras et al.,
1999; Elford et al., 2000; Soyka, 2000; Skotte
et al., 2002; Schibye et al., 2003; Garg, 2006). In
these studies, for example, other types of patient
transfers were analyzed or the transfers were limited
to ‘best-case scenarios’ with low-weight patients or
advantageous posture of the nursing person. In other
cases, inertial effects due to patient-and-HCW
movements or horizontal action force components
were disregarded.
A systematic overview particularly to studies with
emphasis placed on ‘patient transfer’ was provided
by Hignett (2003) and Hignett and Crumpton (2007)
and Hignett et al. (2004). The authors consider both
different transfer techniques and intervention possibilities, which are able to reduce the load for the nursing stuff. In this context, Freitag et al. (2007)
particularly analysed the loads of HCWs caused by
posture during whole shifts. In contrast to the procedure of quantitative determination of short-term mechanical load—aimed to the assessment of an acute
overloading risk for lumbar-spine elements, presented
here—‘dose-oriented’ procedures are well suited for
the description of the cumulative load over the working day or the entire working life by considering the
correlation of loading duration, frequency, and intention, in particular. The assessment of such dose values
take into ancient criteria based on epidemiological
studies about the correlation of long-term mechanical
load and the development of lumbar degenerative diseases (Seidler et al., 2009).
Conclusions
In the presented study, the lumbar load of HCWs
during manual patient handling is described mostly
932
A. Theilmeier et al.
on basis of the force and their components at the disc
L5–S1. The selected ‘3D procedure’ also makes
statements possible about the load aspects regarding
bending and torsional components of force for the
lower spine, which are acting simultaneously as the
forces and require special attention regarding the development of overload-induced disease. In conclusion, aiming an appropriate indication of lumbar
load during patient transfer, the described example
demonstrates the necessity of a time-variant postureand-force determination and a sophisticated biomechanical model. This methodology, however, is
accompanied by a high level of personnel effort for
data gathering and evaluation, and therefore, it is less
well suited for larger subject groups. A detailed
analysis of the subject’s task execution with the help
of the posture-and-force data and a careful selection
of typical tasks from the pool of the totally collected
data could reduce the disadvantages of having a low
sample size as in this laboratory study.
Forecast
As the analyses of 160 representative transfer
actions studied in ‘the Third Dortmund Lumbar
Load Study’ have shown patient transfer activities
may result in intensive lumbar load for the HCWs,
which in many cases exceed recommended lumbarload limits, such as NIOSH’s (1981) action limit or
the age-and-gender specific ‘Dortmund Recommendations’ (Jäger et al., 2001b) for the assessment of
manual handling activities with respect to potential
biomechanical overload of the spinal structures in
the lumbar region. Work evaluation with the obtained data will allow the derivation of preventive
work design measures: the testing of optimized
transfer techniques and of small aids (e.g. sliding
mat or glide board in order to lower the friction between the patient and the bed-or-chair surface) will
presumably show that lumbar load can furthermore
be lowered in order to achieve health prevention in
the everyday working life of healthcare workers.
FUNDING
Statutory Accident and Health Insurance Institution for Health Services and Welfare Care
(Hamburg, Germany).
Acknowledgements—Statutory Accident and Health Insurance
Institution for Health Services and Welfare Care (Hamburg,
Germany) for financial funding; Norbert Wortmann and Stefan
Kuhn (BGW Hamburg / Mainz) for substantial support and
helpful advice; and Barbara-Beate Beck and Beate Wiedmann
(Forum fBB, Hamburg, Germany) for competent and helpful
support and for serving as subjects.
REFERENCES
BKK-Bundesverband der Betriebskrankenkassen. (2008) Seelische Krankheiten prägen das Krankheitsgeschehen. Essen,
Germany: Bundesverband der Betriebskrankenkassen.
Deuretzbacher G, Rehder U. (1997) Messung, Modellierung
und Simulation von Bauarbeitertätigkeiten. Berlin, Germany:
Springer-Verlag.
Elford W, Straker L, Strauss G. (2000) Patient handling with
and without slings: an analysis of the risk of injury to the
lumbar spine. Appl Ergon; 31: 185–200.
European Communities. (2002) European social statistics,
accidents at work and work-related health problems. Luxembourg: European Communities.
Freitag S, Ellegast R, Dulon M et al. (2007) Quantitative measurement of stressful trunk postures in nursing professions.
Ann Occup Hyg; 51: 385–95.
Garg A. (2006) Prevention of injuries in nursing homes and
hospitals. In: Pikaar RN, Koningsveld EAP, Settels PJM,
editors. IEA 2006. Proceedings of the 16th World Congress
on Ergonomics Meeting diversity in ergonomics. Amsterdam, Netherlands: Elsevier.
Garg A, Owen B, Beller B et al. (1991) A biomechanical and
ergonomic evaluation of patient transferring tasks: bed
to wheelchair and wheelchair to bed. Ergonomics; 34:
289–312.
Hignett S. (2003) Intervention strategies to reduce musculoskeletal injuries associated with handling patients: a systematic review. Occup Environ Med; 60: e6.
Hignett S, Crumpton E. (2007) Competency based training for
patient handling. Appl Ergon; 38: 7–17.
Hignett S, Crumpton E, Ruszala S et al. (2004) Evidencebased patient handling tasks, equipment and interventions.
London: Routledge, Taylor & Francis Group.
Hofmann F, Michaelis M, Siegel A et al. (1995) Bandscheibenbedingte Erkrankungen der Wirbelsäule—Untersuchungen zur Frage der beruflichen Verursachung. In
Wolter D, Seide K, editors. Berufskrankheit 2108, Kausalität und Abgrenzungskriterien. Berlin, Germany:
Springer. pp. 47–61.
Hofmann L, Korn M. (2001) Wirbelsäulenbeschwerden bei
Überkopfarbeit—Untersuchungen an Montagewerkern in
der Pkw-Endmontage. In: Drexler H, Ch. Broding H, editors. Paper presented at the 41. Jahrestagung der Deutschen
Gesellschaft für Arbeitsmedizin. Erlangen, Germany:
DGAUM.
Jäger M, Jordan C, Theilmeier A et al. (2007) Biomechanically substantiated reduction of lumbar load for healthcare workers during patient handling. Paper presented at
the ISSA. Athens, Greece: International Social Security
Association.
Jäger M, Luttmann A, Göllner R et al. (2001a) The
Dortmunder—biomechanical model for quantification and
assessment of the load on the lumbar spine. Paper presented
at the SAE Digital Human Modeling. Arlington, VA: Society
of Automotive Engineers. Available at http://www.sae.org/
technical/papers/2001-01-2085. Accessed 1 September 2010.
Jäger M, Luttmann A, Göllner R. (2001b) Analysis of lumbar
ultimate compressive strength for deriving recommended
lumbar-load limits. Paper presented at the Int. Soc. Biomechanics. Zurich, Switzerland: Int. Society of Biomechanics.
Jordan C, Theilmeier A, Luttmann A et al. (2003) Optoelectronic posture recording during patient transfer for determining lumbar load. Paper presented at the quality of work
and products in enterprises of the future. Stuttgart, Germany:
Ergonomia Verlag.
Measurement of action forces and posture to determine the lumbar load of HCWs
Jordan C, Theilmeier A, Luttmann A et al. (2006) Lumbarload analysis for health-care workers during patient transfer
activities. In: Pikaar RN, Koningsveld EAP, Settels PJM,
editors. IEA 2006. Proceedings of the 16th World Congress
on Ergonomics Meeting diversity in ergonomics. Amsterdam, Netherlands: Elsevier.
Luttmann A, Jäger M, Laurig W et al. (1988) Orthopaedic diseases among transport workers. Int Arch Occup Environ
Health; 61: 197–205.
Marras WS, Davis KG, Kirking BC et al. (1999) A comprehensive analysis of low-back disorder risk and spinal loading
during the transferring and repositioning of patients using
different techniques. Ergonomics; 42: 904–26.
Mitchell T, O’Sullivan PB, Smith A et al. (2009) Biopsychosocial
factors are associated with low back pain in female nursing students: a cross-sectional study. Int J Nurs Stud; 46: 678–88.
Morlock M, Hansen I, Bonin V. (1997) Statistische Untersuchung ausgewählter Aspekte der Begutachtung für BK
2108 und biomechanische Überprüfung des Erfassungsbogens EBO 2108 des technischen Aufsichtsdienst. Germany:
TUHH, AB Biomechanik.
NIOSH. (1981) Work practices guide for manual lifting.
Cincinnati, OH: National Institute of Occupational Safety
and Health, NIOSH.
Owen BD, Garg A, Jensen RC. (1992) Four methods for identification of most back-stressing tasks performed by nursing
assistants in nursing homes. Int J Ind Ergon; 9: 213–20.
Riihimäki H, Wickström G, Hänninen K et al. (1989) Radiographically detectable lumbar degenerative changes as risk
indicators of back pain, a cross-sectional epidemiologic
study of concrete reinforcement workers and house painters.
Scand J Work Environ Health; 15: 280–5.
Schibye B, Faber Hansen A, Hye-Knudsen CT et al. (2003)
Biomechanical analysis of the effect of changing patienthandling technique. Appl Ergon; 34: 115–23.
Seidler A, Bergmann A, Jäger M et al. (2009) Cumulative occupational lumbar load and lumbar disc disease—results of
a German multi-center case-control study (EPILIFT). BMC
Musculoskelet Disord; 10. doi:10.1186/1471-2474-10-48.
933
Seidler A, Bolm-Audorff U, Siol T et al. (2003) Occupational
risk factors for symptomatic lumbar disc herniation; a casecontrol study. Occup Environ Med; 60: 821–30.
Skotte J, Essendrop M, Faber Hansen A et al. (2002) A dynamic 3D biomechanical evaluation of the load on the low
back during different patient-handling tasks. J Biomech;
35: 1357–66.
Soyka M. (2000) Rückengerechter Patiententransfer in der
Kranken-und Altenpflege. Bern, Switzerland: Hans Huber.
Theilmeier A. (2006) Erfassung von zeitvarianten Aktionskräften zur Erhebung der mechanischen Wirbelsäulenbelastung bei ausgewählten beruflichen Tätigkeiten (Vol. Nr.
1106). Düsseldorf, Germany: VDI-Verlag.
Theilmeier A, Jordan C, Jäger M et al. (2003) Measurement of
exerted forces during patient transfer for determining
lumbar load. Paper presented at the quality of work and
products in enterprises of the future. Dortmund, Germany:
GfA - Gesellschaft für Arbeitswissenschaft.
Theilmeier A, Jordan C, Luttmann A et al. (2006) Measurement of exerted forces for determining nurses’ lumbar load
during patient transfers. Paper presented at the Meeting
Diversity in Ergonomics. Maastricht, Netherlands: International Ergonomics Association (IEA).
Theilmeier A, Jordan C, Wortmann C et al. (2010) Prevention
of lumbar overload for health-care workers during patienttransfer activities. Saf Sci Monit; 14: 1–8.
Videman T, Nurminen T, Tola S et al. (1984) Low-back pain
in nurses and some loading factors of work. Spine; 9:
400–4.
Videman T, Ojajärvi A, Riihimäki H et al. (2005) Low back
pain among nurses: a follow-up beginning at entry to the
nursing school. Spine; 30: 2334–41.
Waters TR, Putz-Anderson V, Garg A et al. (1993) Revised
NIOSH equation for the design and evaluation of manual
lifting tasks. Ergonomics; 36: 749–76.
Zweiling K. (1996) Katalogisierung wirbelsäulenbelastender
Tätigkeiten. In: Gens W, editor. Paper presented at the 41.
Int. wiss. Koll. Ilmenau, Germany: Technische Universität
Ilmenau.